Newly synthesized DNA, in E. coli lysogenic for the phage X, was labeled by short pulses of [ In bacteria, at least some of the newly synthesized DNA appears as short chains (Okazaki fragments) that are subsequently joined to long molecules (1-3). Whether such discontinuous replication occurs on both strands or only on one is an open question. Evidence, recently reviewed by Klein and Bonhoeffer (4), has been introduced for both possibilities. In order to resolve this question, in the case of Escherichia coli replication, we have studied the molecular polarity of newly synthesized DNA at a given point on the E. coli chromosome, the prophage lambda. The passive replication (5) of this prophage by the chromosomal machinery occurs in a known direction (6), the polarity of integration is known (7,8) and the DNA strands can be easily separated (9). These properties have allowed us to study, by DNA -DNA hybridization, the size distribution and the molecular polarity of the DNA made during exposure to short pulses of tritiated thymidine.EXPERIMENTAL DESIGN Fig. 1 shows two possible models for the manner in which the new DNA strands are formed (1). In model A, (continuous-discontinuous replication) synthesis occurs continuously in the 5' -> 3' direction by addition of nucleotides, to long DNA chains, at the 3' end. Replication in the 3' --5' direction occurs by synthesis of short chains (5' --3', i.e., backward relative to the movement of the replication fork) that are subsequently joined to long DNA.Abbreviation: SSC, standard saline-citrate solution (0.15 M sodium chloride-0.015 M sodium citrate, pH 7); 6 X SSC means that the concentration of the solution used is 6 times that of the standard saline-citrate solution. In model B (discontinuous replication) both strands are synthesized first as short chains then are joined to long DNA.We have isolated the radioactive short fragments made during a short pulse of radioactive thymidine in a strain lysogenic for X, and hybridized these fragments with the separated strands of the phage DNA. Model B predicts that radioactive fragments will hybridize to both strands of phage X, whereas model A predicts that fragments will hybridize only to strand I of X and long DNA will hybridize to strand r.We have found that replication occurs according to model A in our polA + strain. In a polA -strain two size classes of fragments are formed: the smaller class, 10 S, corresponds to the fragments found in the polA + strains; the other class, 12-50 S, corresponds to the strand that is elongated continuously in polA + strains.
MATERIALS AND METHODSBacterial and Bacteriophage Strains. The E. coli Ki2 strain CB 0129 (F-Thy-Leu-BI-) was used. Its replication pat-